摘要
在不重构出原始MPSK信号的前提下进行多元假设,推导出不同假设下压缩样值的最大似然函数,提出一种压缩域最大似然调制识别方法。在压缩感知(CS)框架下,压缩域最大似然调制识别方法(compressive maximum-likelihood,CML)使用远小于传统的基于奈奎斯特采样的最大似然调制识别方法(traditional maximum-likelihood,TML)所需的采样个数实现MPSK信号的调制识别。同时给出了在压缩域进行调制识别的标准,并分析了CML算法的识别的性能。最后给出了该方法的性能仿真。仿真表明CML算法的运算复杂度更低,速度更快,而识别性能仅略低于TML算法。压缩比为1/2,信噪比为-8d B时,CML算法对BPSK信号的识别率达到95%。
Abstract： This paper proposes the Maximum-Likelihood Classification for MPSK signals with com- pressive samplings. Under the compressive sensing （CS） frame, the compressive maximum-likeli- hood （CML） classifier provided in this paper tries to recognize the MPSK signals using far fewer samplings than the traditional maximum-likelihood （TML） classifier which is under the frame of Nyquist sampling needs. This paper presents the criterion of classification and the classification per-formanee analysis. Finally, several numerical simulations are provided and the results indicates that compared with TML, the CML classifier has far lower complexity and slightly lower performance. When the compression ratio is 1/2, and the signal-noise ratio is -8dB, the recognition ratio reaches 95%.
出处
《信息工程大学学报》
2015年第5期579-583,共5页
Journal of Information Engineering University
关键词
压缩感知
调制识别
MPSK
compressive sensing
modulation classification
MPSK
